Evolving descriptive text of mental content from human brain activity
ggm
39 points
37 comments
March 02, 2026
Related Discussions
Found 5 related stories in 52.9ms across 3,471 title embeddings via pgvector HNSW
- Thinking Fast, Slow, and Artificial: How AI Is Reshaping Human Reasoning Anon84 · 116 pts · March 21, 2026 · 54% similar
- Mathematical methods and human thought in the age of AI zaikunzhang · 201 pts · March 30, 2026 · 52% similar
- Artificial intelligence-associated delusions and large language models beardyw · 12 pts · March 14, 2026 · 48% similar
- When Using AI Leads to "Brain Fry" dracula_x · 18 pts · March 06, 2026 · 47% similar
- AI won't fix your family drama, might help you hear what they're trying to say robaato · 12 pts · March 19, 2026 · 47% similar
Discussion Highlights (8 comments)
vlovich123
Prediction: even if this requires surgery, unlocking inner thought will be used in criminal proceedings to establish guilt or attempt to be used to prove innocence. It will definitely be used unethically in military/intelligence interrogations until the law catches up.
dr_dshiv
I have a PhD student working on EEG audio decoding. We are presently focused on a simpler subtopic: the detection of consonance and dissonance in the brain as it listens to music.
ksaj
They don't seem to mention if it is elective. An all or nothing mechanism might spell out words that the patient really didn't intend on others seeing (like "Ugh, that guy again! I can't stand the way he...") It is pretty difficult to control your inner dialog against spontaneous and triggered thoughts.
trocado
"Mental content" seems way to broad for what is rather the sensorimotor part of speech.
jml7c5
As I understand it, the big challenge with brain electrodes is that because they are implanted in a big jiggly piece of jelly, they shift out of position and/or cause localized scarring. The practical effect is that the brain-electrode interface "wears out" after a while, and you can't get useful data. Has this been solved, or are implants still temporary?
voidUpdate
> "It wasn't perfect, but 60% of the words were judged intelligible by testers" I don't understand this part. Are they trying to pull the audio of the words out of the brain or something? I'd think it would be easier to use a dictionary of words, and use some machine learning to try and pull out the most likely next word from the brain activity, in which case 100% of the words would be intelligible
YeGoblynQueenne
>> The answer to whether the tech could identify inner speech was a tentative "yes". For a task involving imagining a sentence, the researchers were able to achieve an accuracy rate of up to 74% in real time. For the tasks designed to prompt spontaneous inner speech, accuracy was reduced but still above chance. Did a number go missing from this sentence? One accuracy rate was "74%", the other was "reduced but still above chance". Why leave things vague? All that accomplishes is that it makes me distrust the factuality of the article.
vivzkestrel
this is like the landline telephone of brain reading tech, imagine what happens when we get an iphone 16 max for the same that works without implants